Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
#qdxm QDX-M - new single-band version of QDX available
#qdxm
|
Hi all I have launched a single?band version of QDX called QDX-M. The 'M' is for monoband. The kit price (and assembly fee if you choose that option) is slightly?less than the price of the? normal multi-band QDX. This reflects the (disappointingly) slight decrease in production costs. Price: $65 (assembly option: $40). QDX-M uses the same firmware as QDX, the only hardware difference is that all band-switching components have been deleted. This simplifies the construction but of course makes it mono-band.? So far in testing the 10/11/12m version of the kit I have found that there is a noticeable improvement in Receive performance. Though the high-bands multi-band QDX also has sufficient sensitivity on 10m, the sensitivity does seem to gain several dB on the mono-band QDX version. This is not unexpected, since the signal no longer has to pass through band-switching multiplexers, and the board layout is simplified resulting in less stray/parasitic inductance and capacitance.? The main motivation for this version was to be able to provide QDX operation?on 160m, 630m and 2200m. On these bands, the operating frequency is too low for the PIN diode switching to work properly. So a single-band version of QDX is necessary for these bands. The board has lower density which means we can use larger toroids for the LF bands to fit the necessary number of turns for the higher inductance.? I will be working on production of the LF versions over the next week or two. Eventually I aim to provide a single-band version for any band from 2200m to 10m.? Initially I only have 10, 11 and 12m versions available. These are in fact all the SAME version, and so this version will work on 3 bands :-)? the only difference, when configured, is the position of the vertical band indicator line on the RF sweep screen in the terminal configuration utilities.? The QDX-M firmware displays only a single entry in the Band Configuration screen. The single entry is configured automatically by the firmware for the band you choose to build the kit for, controlled by your choice of jumper wires (install following the instructions in the assembly manual).? 73 Hans G0UPL |
|
On Thu, Jan 12, 2023 at 10:13 AM, Hans Summers wrote:
Hans, Great rationale for this option.....optimizing the real estate and getting around the pin diode switching. You can also stack smaller diameter LPF toroids for more inductance.? ? 73 Kees K5BCQ?? |
开云体育Just curious, where in the world is 11M digital allowed?Must resist 10M version. ?Must resist. ?Wait, why? ?8*) 73, Willie VP5WS
|
|
Hi Willie 11m JS8 is extremely active in Europe on 27.245 MHz (Channel 25) including some US stations when propagation is good. And at least one Asian on occasionp (me). There's orders of magnitude more JS8 activity on 27.245 than I've ever seen on any amateur band.? Whether 11m digital is allowed in Europe, or it's just that neither the practitioners nor the regulators care in the least whether it's allowed or not, I'm not sure... 73 Hans G0UPL http://qrp-labs.com https://www.buymeacoffee.com/g0upl -------- Original message -------- From: William Smith <w_smith@...> Date: Thu, Jan 12, 2023, 9:21 PM To: [email protected] Subject: Re: [QRPLabs] #qdxm QDX-M - new single-band version of QDX available
|
|
I was wondering whether you could silkscreen the band on the front ...
toggle quoted message
Show quoted text
or maybe on the top as well. 73 Jim N6OTQ On Thu, Jan 12, 2023 at 12:42 PM Hans Summers <hans.summers@...> wrote:
|
|
Hi Jim > I was wondering whether you could silkscreen the band on the? > front ... or maybe on the top as well. You were wondering if **I** could do that? No... I don't have facilities for that sort of thing. The PCB silkscreening is done by the PCB manufacturer. The enclosure manufacturer does the box printing (which is the higher quality laser etching, not silkscreen printed). In neither case could I feasibly create multiple versions, and make some estimate of how many of each band would be wanted... I am shivering uncontrollably, just at the thought...? 73 Hans G0UPL On Thu, Jan 12, 2023 at 10:53 PM Jim Strohm <jim.strohm@...> wrote: I was wondering whether you could silkscreen the band on the front ... |
|
Thanks Tatu I think I missed some setting, which I have now managed to fix. 73 Hans G0UPL On Fri, Jan 13, 2023 at 12:04 PM Tatu Wikman <tatu.wikman@...> wrote: On Thu, Jan 12, 2023 at 06:13 PM, Hans Summers wrote: |
|
Hi Hans,
nice to hear about the QDX mono band version. "So far in testing the 10/11/12m version of the kit I have found that there is a noticeable improvement in Receive performance. Though the high-bands multi-band QDX also has sufficient sensitivity on 10m, the sensitivity does seem to gain several dB on the mono-band QDX version. This is not unexpected, since the signal no longer has to pass through band-switching multiplexers, and the board layout is simplified resulting in less stray/parasitic inductance and capacitance. " I wonder if you know where are the extra dBs on RX coming from, the BPF mpx or the RX/TX LPF switching? I want to try to change my rev.3 first QDX which is for lower bands into a 20m mono (or 20m/17m).? 73 Bojan S53DZ |
|
Hello Bojan I don't know exactly where the dB improvement comes from. And I didn't make a very scientific measurement using my lab equipment, which takes some time to set up. Just a rough estimate from the SNR bar in JS8call. I previously concluded, at least on an unoccupied frequency (no signals, no QRM, only band noise), this SNR measurement in JS8call (or equivalently WSJT-X) correlated well with more rigorous measurements. Furthermore there is of course some build and component tolerance variation from one QDX to another.? Factors that could be involved in the dB improvement... To unknown degrees: 1. No PIN diode losses 2. No losses in the band switching '3253 MUX chip 3. No parasitic inductance and/or capacitance in the unused tap winnings of the L12 BPF inductor.? 4. Shorter PCB traces everywhere due to not needing all the band switching etc, also leading to less stray inductance and capacitance? 5. Use of T37-10 inductors could be improving the situation at the high frequencies compared to T37-6 in the LPF. 6. The T50-6 in the BPF could be less lossy than the T50-2 at 28MHz 7. The single BPF winding spread around the whole L12 core could have better immunity to noise pickup compared to the partial winding on a multiband unit, which is effectively like a solenoidal inductor and has less self-shielding properties All these little pieces of a dB add up... But how the various factors contribute is another question. I sense one could spends months or years studying it. I doubt that on 20m you could use a few dB lower noise floor, the multiband QDX is already very sensitive. The dB improvement may also be not so significant at lower frequencies. Who knows. With my antenna and location (the OCFD antenna not being unreasonable, and I think the location not being a particularly noisy one) the high bands QDX sensitivity is plenty. I don't see that I can use any extra sensitivity available on 10m. With that observation and the fact these.narrow band Digi modes are high SNR inherently anyway.... I'm not sure how many people would be able to use the lower noise floor to advantage anyway.? But it's an interesting observation and one that might hold tantalizing prospects of the door to 6m being less tightly locked.? 73 Hans G0UPL http://qrp-labs.com https://www.buymeacoffee.com/g0upl -------- Original message -------- From: "bojan.naglic" <bojan.naglic@...> Date: Fri, Jan 13, 2023, 4:42 PM To: [email protected] Subject: Re: [QRPLabs] #qdxm QDX-M - new single-band version of QDX available Hi Hans, |
|
Hi Hans. Spanish legal permission for CB are 40 channels, from 26.960 to 27.410 Mhz, with max. power of 4W for AM-FM and 12W for SSB (USB-LSB). But is normal to have more power or frequencies because people want to have better contacts. -- Phill EA5JHA |
|
Hello Phil Yes, 11m is a CB band. Not amateur.? Rules vary according to jurisdiction.? As always, on any radio frequency, the operator should ensure he is compliant with local regulations and his license conditions.? 73 Hans G0UPL http://qrp-labs.com https://www.buymeacoffee.com/g0upl -------- Original message -------- From: "Phill via groups.io" <rocinante2021@...> Date: Fri, Jan 13, 2023, 8:14 PM To: [email protected] Subject: Re: [QRPLabs] #qdxm QDX-M - new single-band version of QDX available
|
|
Hi Hans,
The QDX-M seems a great next step - the idea of a cleaner and more efficient build is very appealing especially at the frequency extremes. I would love to know more about exactly what the obstacles to 6m are - I assume mainly physical tolerances on capacitance and inductance?. I'm only just scratching the surface of my QDX 80m-20m v4, but a single-band 6m version would be VERY difficult to resist. The magic band has always seemed exactly that, but it also has a strong pull of nostalgia for me as the very first antennas I ever built were for TVDXing the old 405-line band 1 signals in the early 80s. I don't think I've seen you comment on 6m much before, apologies if I have missed something in searching. Echoing some other threads I wonder whether it would be possible to assemble a canonical windings and capacitance table for the BPFs and LPFs to allow people to custom-build variations of the multiband versions - e.g. build a 60/40, 30/20, 17/15 version and set the band config accordingly. I'm not suggesting offering kits like that - an inventory nightmare - but to find their own components if necessary. You'd obviously be limited to the 3 LPF paths (or 1 for QDX-M) and I assume the bands would have to be contiguous to keep the BPF sane. Otherwise is it just a sliding scale to choose from? Just thinking aloud rather than proposing anything new - I know how easy it is to add ideas when it's not your list to deliver! 73, Henry, M7FYI |
|
Hello Henry The problem with 6m isn't tolerance of components (inductors and capacitors). In some ways the low inductance makes it easier to adjust compared to the bottom end if the frequency extremes when you have a toroid squeezed full of turns and have no adjustment capability. The main problem for 6m is the speed and noise of the FST3253. The switches in the FST3253 aren't perfect, they don't have zero on resistance, infinite off resistance, and zero switching time. The switching time I think is the main parameter. Simply at very high frequencies there isn't much "on" and "off" time compared to the "in-between" time and this means effectively a higher insertion loss.? At the same time the higher frequency bands have lower ionospheric noise and higher sensitivity is required. So the sensitivity requirement is getting more demanding as the frequency increases, but the receiver is getting less sensitive. Quite likely an RF amplifier would be needed ahead of the detector, for a 6m QDX. You can only go so far with that too... All I'm saying with the 10m QDX-M is that it appears to be very sensitive and this gives, perhaps, a slightly less dismal outlook for the possibility of some day managing a 6m version.? On the other question... I haven't got time in the next few months to experiment with the bandpass and low pass filters for arbitrary combinations of QDX bands and publish a guide. It generally isn't a trivial exercise. The LPFs operate independently. If you life the 20/30m LPF you can re-use it in your own band combo. Etc. But the bandpass is a much more complex issue and really needs experiment to determine the number of turns, which will change for any particular combination of bands. Normal turns calculators for inductors assume a winding uniformly spaced around the core. However for a multi-band QDX that is the case. The winding for the highest frequency covers only a small proportion of the core so the turns are squeezed together and the inductance is therefore higher than it would be for a uniformly spaced set of windings predicted by a calculator. That would change from set of band combinations to another.? If anyone has anything to add or experiments with different bands then can share the results here I'd happily published modifications pages on the QRP Labs website.? 73 Hans G0UPL http://qrp-labs.com https://www.buymeacoffee.com/g0upl -------- Original message -------- From: Henry Johnson <henrybassoon@...> Date: Sat, Jan 14, 2023, 3:56 PM To: [email protected] Subject: Re: [QRPLabs] #qdxm QDX-M - new single-band version of QDX available Hi Hans, |
|
Hi Hans,
Thanks for taking the time for the detailed reply, that highlights a problem I wasn't aware of - section 3.6 of the manual makes a lot more sense in that light and points some more things for me to look into. I suspected the BPF would be non-trivial, I saw one of NA5Y's videos the other day discussing non-uniform spacing: https://youtu.be/raZNgQZ5a-0?t=591 It would be interesting to try to put together a QDX-M with demountable LCs to experiment, but that's definitely something for another day. Good luck with the rest of the QDX-M rollout on the other bands. 73, Henry, M7FYI |
|
On 15/01/2023 09:13, Henry Johnson wrote:
It would be interesting to try to put together a QDX-M with demountable LCs to experimentThe QDX is a super small rig. But I've thought for some time the QDX MAXI would be good. Maybe more of a barebone board with sockets for BPFs and LPFs. Space for on-board regulators, preamp maybe. 73 Alan G4ZFQ |
|
Hi Hans! Thanks again for everything you do!
Just a thought or two to add to your "mulling pot". I hope they add something to mix: Band Pass FIlters In my experience (with my own "radiohat" design and others), the problem with the balanced Tayloe mixer is its strong response to harmonics (especially odd ones). Subharmonics seem to be less of a problem until the signal level reaches one that begins affecting the operating point of the devices. Note however that I have not taken the time to take accurate subharmonic measurements. This is just an impression based on a lot of operational testing. I've been having good luck in my experimental transceiver designs (which make it into the field during the summer) with a single band pass filter ahead of the mixer to reject LF and MW broadcasters on the low end and FM broadcast stations on the high end. The usual transmitter LPF output filter is rejects the HF harmonic multiples. Obviously this is not a perfect solution, and probably only works now that high power SW broadcasting is in decline. A band pass filter is still desirable but these days I have more trouble from strong signals in the same amateur band than from ones outside the band. When I was playing with modifying the QDX, I replaced the BPF with a simple capacitor of low enough value to prevent MW overloading from occurring - this worked well here to finish of the summer's park expeditions, but is clearly not a solution. I had meant to try a 5 element HPF instead, but never got around to it. I did NOT try to bypass the 3253 (but it would probably have helped). In my own experimental designs, I usually precede the mixer with a 7th order Butterworth for 2-32 mhz. Even with SMD parts, the loss can be kept low and the out-of-band rejection high. With miniature axial inductors, it can be quite low. 6 Meters With regard to higher frequencies, the switch performance of the parts is definitely a problem as you go up in frequency. The switch delays become erratic and vary a lot from part to part. Eventually the width of the sampling time becomes short enough to increase the loss significantly. I discovered this when I was trying to find parts with better linearity for my transmitters. The 3253 family parts are really quite non-linear. I never did find any easily available affordable linear switching parts that worked above 20 Mhz, however. Hope this contributes something...?best regards, M |
|
Hi all,
I'm not an RF engineer - as will very quickly become apparent - I freely admit that I really don't know what I'm talking about here. But.. this issue of the quadrature sample detecting at higher frequencies has been buzzing around my head for the last few days and I need to get it out. The analogy I see of the problem at hand is that in having active QSD switching you have a "reciprocating engine" which is hitting a revs upper bound - ideally one would want to move to something with "free rotation to be able to spin faster"; from piston to jet. As I understand it (warning klaxon) the current multiplexer implementation is essentially arbitrary - i.e. capable of going to any switch position from any other as commanded - whereas the requirement here is for a fixed rotary sequence with the critical factor being phase lock. Early PAL colour decoder delay lines, RC timing, transistor chains, and voltage control all come to mind, but things get very messy very quickly. I had wondered if simple logic hardware might provide a solution - something like a 4-bit shift register in a cascade or even flywheel loop configuration turning on fast transistors; but a random sample indicates those clocks max out even before the multiplexers. High speed logic gates in flip-flop or cycling through a simple truth table? PIO bit-banging? At 50MHz x4 even microcontroller clocks look slow... Hmmm. I can't help but think that there "should" be a demod complement to a diode ring mixer or even bridge rectifier; it would be really elegant if the issue could be solved by 4 varicaps acting as both the switch *and* the integration capacitor. I've got Han's "poor man's PIN diode" alternative-use philosophy stuck in my head, in digging around I found some notes he published a few years back on "poor man's varicaps" (http://www.hanssummers.com/varicap/) - it would be very satisfying if a high-speed simple QSD could be implemented by a quadrant of FETs , or even LEDs - "no moving parts" - but that's very wishful thinking indeed. Anyway, very happy to put right on all the misunderstandings above - I get it, if it was that easy it would be have been done - but just sharing random thoughts in the faint hope that might trigger a line of enquiry, but otherwise I'll get my coat... 73 Henry M7FYI |
|
On Thu, Jan 12, 2023 at 10:13 AM, Hans Summers wrote:
The QDX-M firmware displays only a single entry in the Band Configuration screen. The single entry is configured automatically by the firmware for the band you choose to build the kit for, controlled by your choice of jumper wires (install following the instructions in the assembly manual).? If one is to build a QDX-M for a band lower than 20m how should they set up the jumpers? Or does it not matter, as you can change it all in the band configuration in the Terminal program? (I'm thinking 160M, which I know the band name needs to be 160 to get the receiving set up properly.? -Nate? N8BTR |
Phill
to navigate to use esc to dismiss